|Publication number||US7087008 B2|
|Application number||US 10/138,543|
|Publication date||Aug 8, 2006|
|Filing date||May 3, 2002|
|Priority date||May 4, 2001|
|Also published as||CA2443819A1, CA2443819C, EP1383572A2, EP1383572A4, EP1383572B1, US7658704, US20030050527, US20050113630, WO2002089902A2, WO2002089902A3|
|Publication number||10138543, 138543, US 7087008 B2, US 7087008B2, US-B2-7087008, US7087008 B2, US7087008B2|
|Inventors||Peter Fox, Jack Lancaster|
|Original Assignee||Board Of Regents, The University Of Texas System|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (32), Non-Patent Citations (26), Referenced by (64), Classifications (6), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application claims priority to the U.S. Provisional Application No. 60/288,670 filed on May 4, 2001 in the name Peter Fox and Jack Lancaster entitled APPARATUS AND METHODS FOR DELIVERY OF TRANSCRANIAL MAGNETIC STIMULATION.
This invention was made with government support under contract number 5R01MH60246-02. The government has certain rights in this invention.
1. Field of the Invention
The present invention is directed to apparatus and methods particularly suitable for precise aiming and delivery of magnetic stimulation, and more specifically, transcranial magnetic stimulation.
2. Description of the Related Art
Transcranial magnetic stimulation (“TMS”)is a means of repetitively stimulating the human brain through an intact scalp and skull, i.e., non-invasively. TMS is delivered by passing a brief (200 microsecond), strong (10,000 volts, 6,000 amps) electrical current through a coil of wire (a TMS stimulator) placed adjacent to the head. The passage of electrical current induces a strong (2 Tesla) magnetic field which, in turn, induces electrical currents in nearby tissues. In the case of nerve cells, if the induced current is sufficiently intense and properly oriented, it will result in synchronized depolarization of a localized group of neurons (i.e., neuronal “firing”). Initially, magnetic stimulation was used only for peripheral nerves, in which instance it is affecting nerve fibers rather than neuronal cell bodies. More recently (Barker et al., 1985), magnetic stimulation has shown to be able to depolarize neurons in the brain. The cellular element of the brain being affected by TMS was assumed, but not proven, to be fibers rather than neuronal cell bodies.
TMS has several present and potential applications, in the domains of basic neuroscience research and of the treatment of brain disorders. Applications for neuroscience research include, for example: imaging brain connectivity (e.g., Fox et al., 1997); establishing inter-regional and inter-hemispheric conduction times (e.g., Meyer et al., 1995); testing the function of specific brain areas by means of transient functional disruptions, so-called “virtual lesions” (e.g., Shipley & Zeki, 1995); and, studying the modification of synaptic efficacy induced by repetitive stimulation, termed LTP (long-term potentiation) and LTD (long-term depression). Potential clinical applications include, for example: pre-operative mapping, e.g., of language related brain areas (Epstein (et al., 1996); testing for neuronal conduction delays due to dysmyelinating disorders; and, treating brain disorders by selective modification (up or down regulation) of the synaptic efficacy of pathways (i.e., by inducing LTP and LTD; Wang, Wang and Scheich, 1996).
At present, TMS delivery is crude. The TMS effector or stimulator (commonly referred to as a “TMS coil”) is a wire-wound coil, typically shaped like a “B.” The B-shaped coil is placed against the scalp and held in place by a human operator. For the primary motor cortex and primary visual cortex (small sections of the total brain surface), proper positioning is established by the elicited response: muscle contractions when stimulating the primary motor cortex; illusory lights (phosphenes) when stimulating the primary visual cortex. In both of these areas, the effects are very sensitive to coil position and orientation.
For brain regions in which proper positioning cannot be determined by the induced effects (i.e., muscle contractions or subjective experience), position is generally determined by reference to a traditional pattern used for placement of EEG electrodes (10/20 system). The 10/20 system is based on scalp/skull landmarks which do not bear a reliable relationship to the functional anatomy of the brain. Further, when using the 10/20 system, there has been no strategy enunciated for determining proper orientation of the coil. Thus, a reliable method for determining the proper position and orientation of TMS coil placement for brain areas lacking immediately observable feedback is needed.
Application of TMS during radionuclide imaging (using positron-emission tomography (“PET”)or single photon emission tomography (“SPECT”)) has two important uses. First, radionuclide imaging can be used to monitor the induced response, determining precise location and quantifying response magnitude. This is extremely important for testing aiming algorithms and for determining the effect of stimulation parameters, such as intensity, rate, duration and the like. Second, an important use of TMS is to map brain connectivity using radionuclide imaging. For both of these applications, hand-held TMS delivery is inappropriate, for at least three reasons. First, hand-held delivery is unsafe, unnecessarily exposing the experimenter to the radiation used for imaging. Second, hand-held delivery is positionally unstable, degrading image quality by small movements of the holder. Third, hand-held delivery is intrinsically inaccurate and imprecise.
Further, current coil designs for delivery of TMS have been mainly intuitive and somewhat crude. Typical coil designs consist of two loop figure eight type coils, for peripheral nerve and brain stimulation, four loop coils for peripheral nerve stimulation, and variations in angles of these. While attention is paid to coil inductance, it is only for simple circuits that this may be easily calculated.
The target field method has been used to produce minimum inductance cylindrical gradient coils for MRI (Turner, 1986) and has been adapted for bi-planar coils (Martens et al., 1991). Minimum power designs have also been presented (Bowtell et al.). However, such design methods have not been applied to the design of magnetic stimulation coils.
Various combinations of circular or rectangular coil shapes have been designed. Figure eight type designs appear to be the most common. Further, B-shaped and slinky-type coils have also been designed (Cadwell; Lin et al.). Sections of toroids (Davey, Epstein, Carbunaru) with magnetically permeable cores are also known, and appear to be an efficient design. However, none of these provide an extremely focused field penetration. Two and four wing coils with a straight section joined with curves for peripheral nerve stimulation have been designed. (Ruohonen et al.) However, these designs are largely limited by intuition.
Thus, fundamental limitations on the utility of TMS for research and treatment include a lack of methods for precise, automated aiming (positioning and orienting) and safe, rigid (i.e., non-human) holding of the TMS stimulator, as well as the poor suitability of present coils for TMS.
The present invention enhances the precision and ease with which TMS may be used for the diagnosis and treatment of neurological and psychiatric disorders and for neuroscience research. In certain embodiments, these benefits may be accomplished via use of specifically shaped TMS stimulators having certain properties. In certain embodiments, a robot, such as a neurosurgical robot, may be used to deliver TMS. The present invention also includes algorithms for treatment planning and treatment delivery, including: algorithms for rapidly modeling the 3-D electric field created in the brain by a TMS coil at any external location; cortical surface modeling (extraction and visualization); scalar product (electric-field vector times cortical-surface vector) computation and visualization; and merging of functional images, structural images and treatment-planning models (surfaces & fields).
The present invention further includes treatment-delivery tools such as frameless registration of head, brain image, and robot; fully automated robotic positioning of the TMS coil; robotic sensing of TMS orientation (about a manually operated tool-rotation axis); hardware extensions including a passive digitizing arm; a TMS tool mount; a passive tool-rotation axis with an orientation sensor; and a general-purpose mobile cart.
Additionally, inventive methods enhance the precision and ease with which TMS may be used for neuroscience research and for the diagnosis and treatment of neurological and psychiatric disorders.
The inventors have determined that the biological efficacy of transcranial magnetic stimulation applied to cerebrum can be estimated at any point as the scalar product of the induced electrical field (E, a vector) and a unit vector aligned parallel to the cortical columns. The unit vector is estimated as a normal (i.e., perpendicular) the cortical surface, as this is the known orientation of cortical columns. The biological efficacy of a TMS E field, then, is calculated using equation 1.
Biological efficacy=|E| cos θ 
where θ is the angle between E and the unit normal vector. This Cortical Column Cosine Aiming Principle (“CCCAP” or “CAPS” or “aiming principle”)is based on the inventors' determination that the cortical column is the biological unit of the brain with the lowest threshold for TMS excitation and the well-established neuro-anatomical principle that the cortical columns are oriented at a right angle to the cortical surface. Thus, maximum biological efficacy of a cortical region of interest occurs where the induced E field is parallel to the direction of the cortical columns, i.e., normal to the cortical surface.
The CCCAP includes the following principles: (1) surface grey matter (the cortex) is preferentially or exclusively activated; axons in the sub-cortical white matter are minimally activated by the TMS-induced E-field, but will conduct action potentials initiated in cortex by TMS; (2) cortical grey matter will be most effectively activated by an E-field oriented parallel to the columnar organization of the cortex; (3) the response magnitude at any cortical location is a function of the magnitude of the E-field parallel to the cortical columns; (4) cortex is preferentially (but not exclusively) activated by orthodromic E-fields, passing from the pial surface, through the soma, to the sub-cortex; antidromic E-fields, passing toward the pial surface will be less effective but not ineffective; and (5) the orientation of the cortical columns is macroscopically estimated as the normal to the true cortical surface. Of the five aiming principles, the first two (particularly the second) are strongly at odds with current opinion and practice in the TMS community. It is to be understood that a true cortical surface is one derived from an anatomical image with sufficient spatial resolution and image contrast to define the cortical-subcortical or cortical-CSF border. Simplified (e.g., spherical) models or generalized models of the cortex which do not define the true cortical surface may not accurately model the orientation of the cortical columns.
Prior art (Brasil-Neto et. al 1992; Mills et al., 1992) has empirically demonstrated that coil orientation computed in the manner just described is optimal for one brain area (the primary motor cortex), but provided a rationale not applicable to brain regions other than primary motor cortex and inconsistent with the above explanation.
In an example embodiment, the CCCAP may be used to allow the cortical excitation effects of TMS to be computed in advance for any position and orientation, thereby enabling computer-aided aiming of TMS. However, in other embodiments, the CCCAP may be used for manual aiming and orienting. The CCCAP may also be used to normalize (correct) observed biological effects for the angle of intersection with the cortex, when aiming is done conventionally (i.e., not with CCCAP) but images are obtained showing the relationship between the TMS coil and the cortical surface.
In an example method, image-guided, computer-aided implementation of the CCCAP for TMS delivery may be effected in accordance with the following steps. First, an imaging stage is performed, in which a high-resolution, anatomical image (e.g., a 3D T1-weighted image) of the subject's head is obtained. Preferably, sufficient grey-white contrast is obtained to allow detailed cortical-surface segmentation. Next, a functional image (either functional MRI or PET, for example) data set is obtained during conditions (task/control pair) which selectively activate the cortical region of interest (e.g., repetitive hand movement to activate the supplementary motor area).
Next, a modeling stage is performed to identify surfaces within the anatomical and functional images, to be used for registration of the functional and anatomical images to one another and, subsequently, to the patient. In the anatomical image (e.g., an anatomical MRI), the scalp surface and the brain's cortical surface are segmented and modeled as polygon-mesh surfaces. The brain-surface is modeled at high-resolution, to provide an accurate, detailed representation of the interface between cerebrospinal fluid (CSF) and cortical grey matter, as this is critical for establishing the orientation of the cortical columns. Anatomical surface extraction and modeling is done in a manner keeping both surface models (scalp and brain) in register with the original image and, thereby, with one another. In the functional image (either PET or fMRI), the brain surface is segmented and modeled as a polygon mesh surface (but with less detail than the model derived from the anatomical image) and the targeted site is identified. The two brain surface models (anatomical-image-derived and functional-image-derived) are co-registered, thereby bringing the target site in register with the two MRI-derived cortical surface models: brain and scalp. This comprises a conjoined functional/anatomical model.
Then, a 3-D model of the TMS-coil's physical surface and the 3-D E-field induced by the TMS coil is created, and this TMS coil-surface/E-field model is superimposed on the conjoined functional/anatomical model. The TMS coil-surface/E-field model is positioned and oriented so as to obtain maximum biological efficacy (as defined by the CCCAP) at the target point, while keeping the coil surface outside the scalp surface. The position and orientation of the TMS coil-surface relative to the conjoined functional/anatomical model may then be stored for subsequent use.
In an example embodiment, this data may be used to perform TMS delivery in accordance with the following. First, the subject is placed in the treatment position and the head immobilized, for example, using a thermoplastic mask (Fox et al., 1985). With the subject in the treatment position, a 3-D digitizer (e.g., a passive arm digitizer) may be used to collect a series of points on the scalp surface. These points are used to create a model of the scalp surface in the subject's current head position. This model is registered to the conjoined functional/anatomical model previously created, using a rapid, surface matching algorithm, such as a convex hull algorithm (Lancaster et al., 1999). The manual digitizer may again be used to collect a set of specific reference points on the surface of the TMS coil, which may be mounted on a multi-joint, calibrated armature, either passive or robotic.
As the optimal position for the TMS to achieve supra-threshold stimulation of the target location has been previously computed, the translations and rotations needed to move from the present position to the desired position are computed. If the TMS is held by a passive armature, movement is executed manually, with the readout of the coordinates of the armature as feedback. If the TMS is held by a robot, movement is executed automatically, after a safe pathway (avoiding contact with the subject or other obstacles) has been computed. When properly positioned and oriented, TMS delivery is effected.
The advantages of this method are several. First, placement is extremely precise (±1 mm), much more precise than is possible using hand-held aiming. Second, placement can be computed in advance, rather than by trial-and-error, as is done with hand-held aiming for areas with a measurable behavioral response. Third, once computed, a position can be precisely re-established for subsequent treatments. Fourth, the entire process is mathematically specified and suitable for computer implementation. Fifth, the procedure is suitable for application to any cortical location, not just brain areas in which aiming accuracy can be confirmed by an overt behavioral response. Sixth, once the necessary brain images are acquired, trajectories for stimulating any number of functional areas can be computed.
In an example embodiment, TMS delivery may be effected by means of a robotic arm. Using a robotic arm to move, aim and hold the TMS provides many benefits. First, a robotic arm allows the TMS coil to be placed and oriented with accuracy and precision of—1 mm (location) and—1° (orientation), far exceeding hand-holding. This degree of accuracy and precision is crucial for implementing a high-precision aiming algorithm. Second, a robotic arm allows the TMS to be positioned automatically, i.e., under computer control. Third, a robotic arm allows the TMS to be positioned rapidly, saving time for the patient, experimentor or clinician. Fourth, a robotic arm can accurately re-establish its position on sequential days, which is needed for experiments and treatments applied over a series of sessions. Fifth, a robotic arm allows the TMS to be moved rapidly from one position to another, allowing treatment of more than one location in a single session. Sixth, a robotic arm allows the TMS to be held immobile for long periods of time, which a person cannot practically or comfortably achieve. Seventh, a robotic arm allows TMS to be delivered during PET imaging, without exposing a human holder to radiation.
The present invention is also directed to an inverse design method to produce a desired electric field profile for transcranial/peripheral nerve magnetic stimulation. Such a method may be used to design a variety of coils. Such coils may include multiple winged coils having a concentrated bundle of wires at the center and smooth arcs splayed therefrom at an increasing distance, maintaining the minimum inductance and/or minimum power dissipation possible for a given field profile. Similarly, other coils may be used to produce a spatial gradient across the nerve, maintaining minimum inductance and/or minimum power dissipation.
Coils may be designed in accordance with the teachings of the present invention to incorporate one or more of the following characteristics which aid in the delivery of transcranial magnetic stimulation. In certain embodiments, a coil may be designed to provide for minimum inductance for a given set of field constraints. Further, a coil may be designed to dissipate a minimum amount of power for a given set of field constraints. Similarly, a coil may be designed to provide for minimum inductance and minimum dissipation for a given set of field constraints. In certain embodiments, a coil may be designed with negative turns to reduce the electric field on a patient. Such negative turns also may be provided in a separate layer to similarly reduce electric field on the patient.
The present invention may be better understood, and its numerous objects, features, and advantages made apparent to those skilled in the art by referencing the accompanying drawings.
The use of the same reference symbols in different drawings indicates similar or identical items.
Aiming is the most fundamental, unresolved issue for scientific and medical uses of TMS. TMS is most often “aimed” merely by observing its behavioral effects. For example, primary motor cortex is frequently identified by adjusting coil position/orientation to achieve a contraction of the desired muscle (e.g., abductor pollicis brevis) at the lowest stimulus voltage. Very recently, image-guided aiming has been introduced (below). Even when image-guided, coil orientation relative to brain anatomy has either been ignored (Paus et al., 1997) or empirically derived (Krings et al., 1997), but it is not based upon a general theory of aiming. Preliminary conjectures as to the mode of interaction between TMS and the brain have been put forward, and emphasize interactions with fibers running parallel to the cortical surface (i.e., horizontal relative to the vertical orientation of the cortical columns). While many accept the horizontal-fiber hypothesis discussed (below), no thorough treatment of its implications for TMS aiming exists. In fact, its predictions do not agree with a growing number of observations. The status and shortcomings of TMS aiming theory and practice are synopsized here, as further background to the present invention.
In a planar electromagnetic coil, the induced E-field parallels the current in the coil, is maximum in the plane of the coil and falls off rapidly with distance from the surface of the coil. Presently known TMS coils are made in two basic geometries: (1) circular (O-shaped, as shown in
A much more focal E-field can be created by placing two circular coils, current flowing in opposite directions (i.e., clockwise and anti-clockwise), next to each other, creating a B-shaped coil. For this coil, the E-field is enhanced where the coils are near each other, because the electric fields from each coil sum. The result is a focal E-field in the center of the coil, oriented parallel to the central region (short-axis) of the B-shaped coil (
Early O-coil studies (Barker et al., 1985; Barker et al., 1987) observed that motor responses could be obtained from either hemisphere by reversing the direction of current flow. They concluded that this was due preferential sensitivity of neurons to orthodromic currents, recognizing that neural structures “are more likely to be stimulated if they are oriented parallel to the electrical field lines” (Barker et al., 1987). Nevertheless, early researchers remained agnostic on the brain location (cortical vs. subcortical; sulcal vs. gyral) and specific neural elements being stimulated. For example, Yeomans (1990) comments that “localization and orientation effects of magnetic stimulation have not been explained neurally” (pg. 141). In a similar vein, Roth et al. (1991) concludes: “Although the use of magnetic stimulation is growing rapidly, the technique has been applied without a complete theoretical understanding of the induced electric field distribution”; and, “none of these studies define a definitive relationship between the electric field distribution in the motor cortex and the resulting transmembrane potential induced in the cortical neurons.”
To this day, many remain agnostic on this matter. As a consequence, TMS position (placement and orientation) follows no coherent theory. Some users position the TMS coil by behavioral optimization (e.g., minimum-threshold muscle twitch); some position carefully, but ignore orientation, using the B-nose convention (discussed below); some orient by reference to prior behavioral-optimization studies; none use an aiming theory.
Day (et al. 1989) were the first to postulate a specific site/mechanism of action for TMS. In the context of a study of primary motor cortex, Day reasoned that “stimuli are likely to activate those neurons nearest the stimulating electrode, that is, those on the convexity of the precentral gyrus”. Day knew that an O-shaped coil, positioned flat against the scalp, produced a circular E-field tangent to the gyral crown and at a 90° angle (horizontal) to the (vertical) columnar organization of the cortex at the gyral crown. Day accepted the long-established principle that “a voltage gradient parallel to the long axis of the neuron is the most favorable” (Day et al., 1989). Consequently, Day postulated that horizontal fibers (interneurons, pyramidal tract axon collaterals and afferent axons from cortical and subcortical sites) at the gyral crown must be the directly activated neural elements, with pyramidal neurons (projecting to the spinal cord) being secondarily activated. Note that Day's inferences were based on the assumption that TMS would only be able to activate the brain region nearest to the coil surface, i.e., the gyrus. However, the data of Fox (et al., 1997), Paus (et al. 1997) and the EXEMPLARY STUDY below using PET imaging to detect TMS effects on the brain, clearly indicate that TMS applied with a B-shaped coil preferentially stimulates sulci, rather than gyri. Thus, the premise of Day's reasoning was incorrect.
Experimental support for the horizontal-fiber hypothesis is notably lacking. Day presented no experimental evidence of horizontal fiber activation. The cortical electrical stimulation literature (which excites cortex by passing applying electric potentials directly to the scalp) has consistently utilized current applied perpendicular to the cortical surface at gyral crown (i.e, in the orientation prescribed by the CCCAP) and never horizontally, as would be predicted by Day's horizontal-fiber hypothesis. In fact, no evidence has ever been provided that horizontal-fiber activation results in pyramidal neuron activation or, more importantly for the present proposal, in columnar excitation comparable to that of physiological activation or vertical electrical cortical stimulation.
Theoretical flaws in the horizontal-fiber hypothesis are readily identified. The horizontal-fiber hypothesis postulates that pyramidal motor neurons, located on the anterior bank of the central sulcus, are activated by means of horizontal fibers on the crown of the pre-central gyrus. This is a distance of 0.5–2.0 centimeters. Yet, horizontal fibers extend only 1–2 mm (Jones and Wise, 1978; Jones, 1981). Further, horizontal fibers are isotropic, extending uniformly in all directions within a plane parallel to the cortical surface. The isotropism (lack of a preferred orientation) of the horizontal fibers should translate into a lack of a preferred orientation for TMS, as the E-field should excite a roughly equal fraction of the total horizontal-fiber population in any orientation. This has been clearly disproven by some of the very studies which espouse the horizontal-fiber hypothesis. Finally, the horizontal fiber hypothesis is weakened by the fact that the predominant horizontal element providing for horizontal interactions is the basket cell, which “should be considered a class of large inhibitory interneuron” (Jones, 1981). Thus, the hypothesis that TMS excites pyramidal cells by means of horizontal fibers is quite unlikely.
Current common practice is to place a B-shaped coil flat against the scalp approximately over the area to be stimulated, as shown in
Hand-held, non-image-guided application of TMS is the norm. This rather casual approach to stimulation position and orientation likely has several contributing factors. First, hand-held aiming is simple to implement and inexpensive. Second, many users of TMS conceive its effects to be rather diffuse. Consequently, they see little need to be highly precise in aiming. Direct evidence that TMS' effects are highly localized only appeared recently, by PET imaging during TMS (Paus et al., 1997; Fox et al., 1997). Third, no general aiming principle had been enunciated to drive greater precision. Fourth, no systems for facilitating precise, image-guided aiming are available.
Paus (et al. 1997) used a modified commercially available passive digitizing arm, namely a VIEWING WAND® (ISG Technologies, Toronto, Canada) to perform a first-generation form of image-guided TMS. TMS was delivered to the average location of the frontal eye fields (FEF), defined stereotactically (i.e., in standardized coordinates) as the mean of eight, previously published group-mean, PET-activation studies (Paus et al., 1996). Cortical surface geometry (sulcal/gyral location and orientation) was not taken into consideration either for placement or for orientation. Orientation was “B-nose”. Anatomical MRI was obtained but was used solely to establish the inverse transformation from the standardized space (Talairach and Tournoux, 1988) to the individual's native image space. No functional markers (e.g., PET during eye movements) were obtained in the subjects either prior to or during the TMS study. The response location was imaged with PET, but was not interpreted relative to group or individual cortical geometry nor was any discussion of an aiming theory included.
Krings (et al. 1997) used a modified SURGICAL ARM® (Radionics, Burlington, Mass.) to stimulate primary motor cortex. The purpose of this study was to compare response locations for TMS and EBS, validating TMS for pre-operative mapping of motor cortex. Position was adjusted to achieve a consistent motor response (thumb twitch). Orientation was based on anatomical MRI, positioning a B-shaped coil with the short axis perpendicular to the central sulcus, in keeping with the recommendations of Brasil-Neto et al. (1992) and Mills et al. (1992). However, Krings offers no aiming theory. The orientation rules which Krings cites (Brasil-Neto et al., 1992; Mills et al., 1992) are specific to primary motor cortex and do not constitute a generalized aiming theory.
Extremely important factors for any TMS aiming system to address are alterations of the induced fields by the stimulation apparatus or the stimulated object. The TMS-induced B field falls off rapidly with the distance from the coil. The induced B field can be altered by nearby ferro-magnetic materials, via the creation of secondarily induced B-fields. The magnitude of a secondarily induced B field is a function of the μ of the metal and the amount of metal; the shape of the secondarily induced B-field is a function of the shape of the object. The secondary B field and the TMS B field will sum, which can distort the net B field in a complex manner. Materials with high μ are metals. Thus, it is desirable to use a robotic arm having low μ. Furthermore, interaction between coil and arm may be minimized by creating distance between the TMS coil and arm. Biological tissues, in general, are paramagnetic, having very small μ; thus, they do not significantly distort the TMS-induced B field.
Unlike the B field, the brain's TMS-induced E field is subject to alterations from biological tissues, as follows. Following the TMS-induced E field, currents flow. At the interfaces of tissues whose conductivities differ, charge accumulates. For the head there are three important conductivity interfaces: 1) air and scalp; 2) scalp and skull (outer table); and 3) skull (inner table) and brain. While minor conductivity differences do exist among the several extracranial tissues (dermis, subdermal fat, galea) and intracranial tissues (i.e., meninges, blood vessels, grey matter, white matter and CSF), these differences are small relative to the three major interfaces above. Soft tissues (scalp and brain) are weak conductors; air and skull are non-conductors (insulators). Charge build up occurs, therefore, at the air-scalp, scalp-skull, and skull-brain boundaries. The shape of the head and skull determines the geometry of the charge build up and, thereby, the precise geometry of the E-field distortion. The greatest change in the TMS-induced E field will occur nearest the accumulated charge. Research continues in this area.
The CAPs predict the results of the two orientation-optimization experiments in the TMS literature (Brasil-Neto et al., 1992; Mills et al., 1992), and the two PET/TMS papers published to date (Paus et al., 1997; Fox et al., 1997), and an exemplary study testing the aiming principles.
The optimal orientation for TMS excitation of primary motor cortex for hand (M1-hand) was simultaneously reported by two independent research teams (Mills et al., 1992; Brasil-Neto et al., 1992). Mills stimulated the left hemisphere; Brasil-Neto stimulated the right. Both groups used the “B-nose” orientation as the reference (0°) orientation, rotating the coil in 45° increments. Both groups found the optimal orientation to be rotated medially 45°, and interpreted this as being perpendicular to the central sulcus, as shown in
Brasil-Neto (1992), on the other hand, postulated that “horizontal interneurons which are aligned perpendicular to the central sulcus are preferentially activated by magnetic stimuli”, but provided no rationale or supporting data. Both Mills and Brasil-Neto cited studies which indicate the existence of horizontal fibers and that some fraction of these fibers are in the appropriate plane. Neither presented a rationale for preferential orientation (i.e., non-isotropism) of horizontal fibers nor a rationale for their selective orientation perpendicular to the central sulcus. Finally, neither Mills nor Brasil-Neto addressed the great distance (1–2 cm) between the presumed site of stimulation (gyral crown) and the location of human motor cortex (anterior bank of central sulcus) nor the inhibitory nature of horizontal interneurons. Thus, both studies are problematic for the horizontal-fiber hypothesis, even though both cited the horizontal-fiber hypothesis as their underlying theoretical construct.
On the other hand, these orientation-effect results are exactly predicted by aiming principles 2 and 3, which predict maximal activation when the E-field is perpendicular to the cortex and a graded response proportionate to the total Ev. When a B-shaped coil is positioned over a sulcus, the vertical component (Ev) of E is maximal when E is perpendicular to the sulcus, as shown in
The TMS/PET experiments of Paus (et al., 1997) and Fox (et al., 1997) provide strong evidence for the columnar aiming principles set forth above. Paus applied TMS with a B-shaped coil to the mean location of the frontal eye fields (FEF). The resulting activation was recorded as an increase in cerebral blood flow using H2 15O PET. For purposes of the aiming principles, Paus' most notable finding was that the induced activation was sulcal (not gyral), lying approximately 3 cm deep to the gyral crown. This is in agreement with the prediction of principle 2 and contradicts the prediction of the horizontal fiber hypothesis.
Fox and colleagues applied TMS with a B-shaped coil to primary motor cortex. The coil was positioned and oriented behaviorally, to elicit a contraction of the abductor pollicis brevis at the lowest possible threshold. The resulting activation was recorded as an increase in cerebral blood flow using H2 15O PET. Data underwent a group-mean analysis to create z-score activation images overlaid on the averaged MRI. Again, the induced activation (at the site of TMS application) was sulcal, lying 2–3 cm deep to the gyral crown. Here, the activation site was the anterior bank of the central sulcus, in agreement with known physiology and histology and clearly supporting principle 2.
The inventors have carried out two studies of aiming principles 2 and 3: one studying E-field intensity (Experiment 1) and a second studying E-field orientation (Experiment 2), as follows:
Experiment 1: By Principle 3, maintaining the E-field at a field orientation but varying the strength of the field should vary the degree of TMS-induced neuronal activity. Eight normal volunteers were imaged by PET while undergoing 3-Hz TMS stimulations across a range of intensities. Stimulation was applied to the Supplementary Motor Area on the medial surface of the right frontal lobe of the cerebral hemisphere. The E-field was oriented perpendicular to the cortical surface, which was also perpendicular to the mid-sagittal plane and exactly 90° rotated from the orientation required by the “B-nose” principle, described above. This stimulation produced strong activation on the medial-bank cortical surface at the location in which the E-field was most perpendicular to the cortical surface, as predicted by Principles 2 and 3. Response magnitude was highly correlated with TMS E-field strength, as predicted by Principle 3.
Experiment 2: By Principles 2 and 3, varying the orientation of the TMS E-field relative to the cerebral cortical surface should modulate response magnitude, with the greatest response being observed when the E-field is perpendicular to the cortical surface. Twelve normal volunteers were imaged by PET while undergoing 3 Hz TMS at fixed intensity but varying orientation: 0°, 30°, 60° and 90° relative to the cortical surface. Stimulation was applied to the foot area of primary motor cortex. As predicted by Principles 2 and 3, the response was located on the bank of the central sulcus, rather than on the gyral crown. Also as predicted by the same principles, response magnitude was highly correlated with TMS E-field orientation, with the greatest response occurring when the field was at 90° to the cortical surface, as shown in
To take advantage of the aiming principles set forth herein, it is desirable to use a TMS stimulator having properties which allow it to generate an E field having a maximum at a clearly defined location, said location being easily placed in close alignment with a subject's head, so that the E field is directed substantially parallel to the cortical column of interest. Furthermore, in an example embodiment, such a stimulator (or a standard TMS stimulator) may be used in connection with a robotic system to permit for accurate delivery of the E field. It is to be understood, however, that the aiming principles set forth herein may be used in connection with known TMS delivery modes, and likewise, the novel stimulator and delivery system may be used in TMS delivery modes without use of the aiming principles set forth herein.
A flat, B-shaped coil, such as is most commonly used for TMS, creates an E-field that is oriented tangential to the scalp surface upon which it is placed. By Aiming Principles 2 and 3, this E-field configuration is optimal for stimulating the banks of sulci in which the sulcus is oriented perpendicular to the scalp surface (e.g., in
In an example embodiment, the power supply may include a high energy capacitor bank, which when discharged provides a high current through the coil portion 120. In an example embodiment, the electrical current may be relatively strong (between about 1000 and 2000 amps) and last for a relatively short period of time (between 50 and 250 microseconds). To deliver such a high intensity, short pulse current, the power supply may include a capacitor bank having a very high storage capacity on the order of 50 microfarads charged to a potential of 2000–6000 volts. Preferably, the capacitor bank is made of high duty material to withstand the current generated. For example, the capacitor have a capacitance of 50 MFD and a voltage rating of 7.5 kV may be made physically larger to extend the lifetime from an estimated 103 pulses to 108 pulses. Furthermore, in an exemplary embodiment, the power supply, conductors, and the TMS stimulator may be water-cooled to reduce operating temperature.
Although shown in
In an example embodiment, the body portion 110 may comprise a cylinder having a diameter between about 10 and 20 centimeters, and a hollow core having a diameter between about 0.5 and 2.0 centimeters.
In an example embodiment, the wire portion 120 extends substantially around the body portion 110 in a toroidal shape. In certain embodiments, the wire used may have a diameter between about 0.1 and 1.0 millimeters.
By use of such a design, when energized by an electric current, the TMS stimulator 100 generates a maximum induced electric field extending along the long axis 130 of the body portion 110.
In an example embodiment, TMS delivery may be effected via use of a robot. More specifically, a robot having six or more degrees of freedom may be used to appropriately position and orient the TMS coil in a precise location and orientation for most effective delivery of induced electric field. Without a sixth degree of freedom, the orientation aspect of the CCCAP described above cannot be implemented. In an example embodiment, a surgical robot having five joints may be modified to add a device to hold the TMS stimulator and provide for a sixth degree of freedom. Alternately, a robot having six degrees of freedom may be used.
The robot may be a medical robot to provide benefits of a robot safe and effective for use with human patients. Such a robot may be controlled by a computing system, such as a PC. Robot tool motion and orientation may be controlled from appropriate software programs. Such programs may include control for robot arm position monitoring, position modeling, plan movement modeling, and movement model execution. Additionally, sensors may be provided so that XYZ position and angles for each of the joints of the robot arm may be obtained in real time. Motion commands may be executed in a coordinate system native to the robot. In an example embodiment, the robot arm may be moved from point to point in a smooth manner using a 3-D mouse, SPACE MOUSE™, or keyboard input. By use of such a robot, maintenance of a pre-specified treatment position may be more precise and accurate than a passive arm or a hand-held device. Furthermore, a robot allows movement through a treatment zone (rather than treatment of a point) and further allows precisely timed treatments of multiple sites in a single session.
Via use of such a robot, TMS may be effective for use in connection with pre-operative mapping, inter-operative monitoring, and treatment of conditions, such as long-term depression, and for clinical activity and functional mapping using TMS and PET. Additionally, by use of a robot, a means is provided to export a generic treatment plan. In other words, a coordinate system may be developed and standardized across different patients, labs, and the like. Furthermore, such use of a robot is particularly useful with regard to a treatment program in which a patient is subjected to TMS over a repeated number of days.
In an example embodiment, a NEUROMATE™ stereotactic assistant system (ISSI, Sacramento, Calif.) was modified for use.
Robotic devices must be used with caution in the presence of persons. In an example embodiment, a robot such as the NEUROMATE™ specifically designed for use by and with humans is desired. Benefits include that it moves slowly, such that it can easily be stopped prior to any collision, loss of electrical power renders it immobile in its current position, i.e., it does not return to a “parking” position or make any sudden motions. Further, such a robot may require a key and a recessed button to actuate, being used only under direct supervision.
A TMS-coil holder was constructed for the TMS coil to adapt to the NEUROMATE™. In the example embodiment, the TMS tool-holder included a tool-rotation axis, as lacking this, changing coil orientation would require establishing a new arm posture, with only a very limited number of orientation angles (at a given location) being possible. In addition, the tool holder serves as a stand-off, creating a distance between the final limb of the robotic arm and the TMS coil. This distance decreases effects of the metal arm on the coil's B and E-fields. In addition, this helps keep the metal arm out of the PET field-of-view, thereby lowering the possibility of PET-image attenuation artifacts. Finally, the tool holder permits orientation sensing and motorized, computer-controlled coil-orienting. It is to be understood, however, that a tool holder need not include a TMS-rotation axis, and that in other embodiments (such as a 6-joint robot), the TMS coil may be adapted directly to a robotic arm.
In the example embodiment, the TMS-coil tool holder was constructed of fiberglass. However, in other embodiments, the tool-holder may be made from a number of other materials, such as acrylic, fiberglass, delrin, or the like. As shown in
In an example embodiment, a precision rotional sensor 250 may be mounted on the end of the proximal (belt-driven) rod. In an example embodiment, the sensor may be a single turn (340°, 10 kΩ) precision resistive potentiometer with +/−1% linear rating. The output of the sensor specifies the orientation of the TMS coil. The sensor's output may be conveyed to TMS delivery software via tool-sensing electrical contacts on the robotic arm tool-mounting plate 310. In neurosurgical applications, these contacts are used to monitor the depth of tool (e.g., drill-bit or wire electrode) passed through a stationary tool holder. In the present application, they may be used to monitor orientation rather than depth. Additionally, a stepping motor may be added, allowing active rotations of the TMS coil.
As discussed above, the NEUROMATE™ was designed as a system for framed stereotaxy. A patient is registered to the NEUROMATE™ by having his (or her) head placed within a metal ring that is rigidly mounted to the NEUROMATE™ by means of a pedestal attached to a long leg, which forms the base/stand of the NEUROMATE™. For TMS delivery, the robotic arm was converted into a frameless stereotactic system, capable of being rapidly registered to a subject in any position within the operating range of the robot arm, including supine in the PET scanner. It is to be understood, however, that the methods and apparatus of the present invention may be used in conjunction with framed stereotaxy.
In an example embodiment, this registration process may use a manual digitizer to collect points from the subject's scalp, which may be registered to the scalp surface modeled from that subject's image, such as obtained from MRI. In other embodiments, the robot may be used in an interactive mode to collect these scalp points.
In an example embodiment, a 5-axis manual digitizer (Microscribe 3D) may be permanently mounted to the robot chassis for use as required. The positional relationship between the digitizer and the TMS probe may be calibrated and thereafter should remain constant, although it is recommended that calibration be verified for each use of the system.
Typical surgical robotic devices come equipped with a stand designed for framed stereotaxy. This stand has a long foot which prevents it from being placed sufficiently close to the PET for use during TMS/PET. Thus, a shallow, wide, weighted, four-wheeled, cart may be used having weight and width to prevent tipping. Preferably, the wheels are retractable, to allow the system to be completely stable when in operation, yet re-positionable to optimize placement of the robotic arm relative to the PET.
As presented above, passive digitizing arms can be used for image-guided TMS-delivery. Both Paus (et al., 1997) and Krings (et al., 1997) used commercially available passive arms designed for frameless stereotaxy to assist with TMS aiming. However, these two systems suffer from significant and virtually identical limitations. Neither can be fixed rigidly in position; that is, both are manual digitizers but neither is a holding device. (Paus mounted the TMS coil to a separate, home-made holding arm; Krings held the TMS coil by hand.) Neither system can digitize or visualize tool orientation, as both are five-joint arms with no capability for rotation about a tool axis. (Paus fixed the coil in the B-nose orientation; Krings hand-held the coil in the orientation recommended by Brasil-Neto [et al., 1992] and Mills [et al., 1992]). Not unexpectedly, neither tool has software for visualizing tool position/orientation vectors. Neither system has any capabilities for modeling the E-field, visualizing E-field lines relative to brain anatomy. Finally, neither system is robotic.
An active (robotic) arm has clear long-term advantages. Ideally, treatment positions are pre-specified, with the arm moving to and maintaining the treatment position. A robotic system can carry out this function more precisely and more accurately than a passive arm. Further, only a robotic system readily combines the tool aiming and tool holding functions. Finally, only a robotic arm allows movement through a treatment zone (rather than treating at a point) or allow precisely timed treatments of multiple sites in a single session.
The PET image, the MR image, the manual digitizer, and the robotic arm robot have different coordinate-reference frames. In an example embodiment, a whole-brain, high-resolution (1 mm3), T1-weighted MRI may be used to create a reference space within which all objects become registered. This image format provides ample anatomical detail for precise registration. Additionally, all images may be standardized by removing differences in orientation and position between patients. In an example embodiment, each patient's reference MRI may be rotated and translated (but not scaled) to the alignment of the Talairach and Tournoux (1988) atlas. This six-parameter spatial normalization (of the reference MRI) may be performed using the previously validated SN software (Lancaster, et. al. 1995). This software may be used to freely rotate and translate a 3-D brain image to position it in a standard pose. The standard pose is with the brain's anterior commissure (AC) located at x, y, z,=0,0,0. The line between the anterior and posterior commissures (AC and PC) adjusted to lie along the y-axis and the inter-hemispheric plane fitted to the y-z plane at x=0. The reference MRI may serve as the reference frame for coordinates from: functional images (PET), the head-surface digitizer; the robotic arm, the TMS coil (held by the arm), the standardized head models; and the TMS-induced B and E fields.
In an example embodiment, the robot may be moved in a user-guided visual-feedback mode for manual positioning and holding of the TMS probe. Preferably, the robot may be moved from point to point in a smooth manner using joy stick-like or keyboard input. For example 3-D mouse/joystick, such as a SPACE MOUSE™ may be used in conjunction with a supervisory PC controller to send motion commands to the robot's on-board controller. Motion commands may be executed relative to: (1) the tool coordinate system referenced to the face of the TMS; (2) a reference MRI (aligned but not scaled to the Talairach & Tournoux  brain); and (3) the global coordinate system native to the robot. In order to implement this control feature, the supervisory controller may use a robot inverse transformation to calculate joint coordinates needed to accomplish an incremental change of position relative to tool coordinates. The calculations may take into account arm configuration and joint rotation limits, and will format the commands to the world space required by the supervisory controller.
In an example embodiment, the communication protocol, position-sensing interface specifications, movement sensing interface specifications and complete kinematic equations for the robot arm may be developed using, for example, Microsoft Visual C++ on a Pentium-level PC.
The protocol used by the supervisory control software to communicate with the robot controller may use RS-232 connections and protocol to send instructions and retrieve supervisor-level data from the controller. Using such protocols, a Visual C++ application may support all motion type commands, including an exclusive stop command and polling commands for status.
A second serial port of the PC may be used to get robot arm position data and status. The communication protocol may be similar to that for the supervisory control software discussed above. A multi-port serial I/O board may be added to the PC to support simultaneous access to both ports from the same computer. Supervisory control software routines may be used to read this port and obtain x-y-z position and angles for each of the six joints of the robot arm. Furthermore, forward and inverse kinematic equations may be used to specify and modify arm pose.
In an example embodiment, the patient's head and the robot may be registered to the MR reference frame for each patient and each TMS session, as accurate positioning of the robot forms the basis for registration of the TMS coil relative to the head. Physical coordinates digitized about the surface of the head may be used to register the patient's head and to calibrate the digitizer relative to the MRI reference frame. As disclosed above, the manual digitizer may be permanently mounted on the robot, enabling direct digitizer-to-robot calibration. Rigid-body coordinate-transformation matrices may be calculated between the robot, the digitizer, and the MR reference frame.
For patient registration, a manual 3-D digitizer (MicroScribe 3DLX) may be used to acquire a “cloud” of distributed points on the patients head in a hat-file format (Pelizzari et al., 1989). An MRI surface (head-file format) with 6000–7000 points uniformly distributed about the head may be extracted from the subject's standardized MR image using, for example, convex hull (CHSN) registration software (Lancaster et al. 1999). A coordinate transformation for calibration (or registration) may be determined by fitting the digitized hat “cloud” (patient) to the detailed head (MRI) surface. The fitting method used may be an iterative least square technique that has been shown to provide excellent results for co-registering PET, CT, and MRI (Lancaster et al., 1997). Patients may be positioned within the PET scanner with head constrained. The CHSN software may be used for fitting the hat to the head surfaces and calculating a head-to-MR transform matrix. The digitizer-to-MR matrix is identical to the head-to-MR matrix, as both are calculated using the manual digitizer and the head surface.
For robot registration, the robot-to-digitizer transform matrix may be calculated using fiducials for which both digitizer and robot coordinates can be readily determined. A pseudoinverse method may be used to calculate the transformation matrix (Castleman, 1996). The robot-to-MR transform matrix may be calculated by concatenating the transform matrices for robot-to-digitizer and digitizer-to-MR.
When the TMS coil is permanently mounted on the robot arm, the TMS-to-robot calibration transform matrix can be calculated. This may be done with the coil in a designated zero-degree orientation relative to the mount. Three parameters (x-y-z position, coil, tool, or z axis orientation, and rotation about coil, tool, or z-axis) are used to fully aim the TMS coil. These may be determined using the manual digitizer to record fiducials on the coil's x- and y-axes. Using these data, equations to calibrate the robot can be determined and robotic coil aiming enabled. This calibration should remain fixed but may be verified periodically. Coil aiming may be extended to the MR reference frame using the robot-to-MR transform matrix. Only position and z-axis orientation are under direct control of the robot. The uncontrolled rotation about the z-axis may be calculated for each robot arm position. Targeted orientations of the coil about its z-axis may be achieved using the manual digitizer and/or the orientation sensor.
In order to overlay PET (functional) and MRI (anatomical) images as part of treatment planning, the PET images may be registered to the MR reference frame for each patient. PET images may be co-registered to standardized MR images using, for example, the convex hull (CHSN) registration software (Lancaster et al., 1999). Convex hull surface is extracted from both MRI and PET images. These models remove concave regions that are well resolved by MRI but poorly resolved by PET. However, the convex brain surface is identical in both imaging modalities and proves and excellent surface for surface matching and, thereby, volume registration. The CHSN software (Lancaster et al., 1999) uses the convex hull of the brain, a simplified surface which is highly similar across imaging modalities, for registration and spatial normalization. An iterative least square method is used to fit the convex hull from PET (hat format) to the subject's reference MRI (head format), allowing only rotations and translations of the hat data (Pelizzari et. al., 1989). This fitting method has very small error: mean RMS errors, measured as distance between the two convex hulls, were reported to be less than 1 mm (Lancaster et al., 1999). A PET-to-MR coordinate transformation matrix may thus be calculated by the CHSN software.
There are four principles that can be used to estimate important aiming parameters of the TMS stimulator without resorting to comprehensive E-field calculations. First, the E field magnitude of figure-8 coil tends to remain strongest at points along the central axis perpendicular to the plane of the coil (z-axis). Second, the E-field magnitude rapidly decreases with increasing distance from the surface of the coil. Third, inside a conductor (e.g. brain) the surface charges reduce the E-field magnitude relative to what its magnitude would be at the same point in air. Four, under certain conditions of symmetry a weak conductor (e.g. the brain) will not change the direction of the E-field on the coil's z-axis. For example, if the coil is placed tangential to the head, at a point of fairly symmetric curvature, this symmetry about the z-axis will tend to cancel distortions that would change the direction of the E-field. These “on-axis” principles, in conjunction with the columnar aiming principles may be used to accurately aim the TMS.
To create a safe and effective treatment plan, the following exemplary steps may be performed. First, a cortical stimulus-delivery site must be selected. Second, at the stimulus-delivery site, the columnar orientation must be determined. Third, TMS coil aiming parameters must be determined. Fourth, a model of the TMS E-field (corrected for head-induced distortions) must be determined. Fifth, models of the TMS coil and patient head surface must be provided for collision avoidance. However, it is to be understood that in other embodiments, more or fewer steps may be performed.
In an example embodiment, these steps may be performed via TMS planning software. Site selection may be done interactively while viewing a high-resolution MR image of the brain in three orthogonal views. The columnar orientation at the treatment site may be determined from the high-resolution MRI by means of cortical-surface extraction, as cortical columns are oriented perpendicular to the cortical surface. Once the treatment site and its columnar orientation are established, the on-axis plan proceeds by determining the TMS coil position for aiming. This may make use of a 3-D front-surface model of the coil and the patient's 3-D head surface model to avoid head contact. The coil position closest to the head in which the on-axis E field parallels the cortical columns at the treatment site may be established. The TMS coil aiming parameters (position, coil z-axis orientation, coil y-axis orientation) may be determined and used to direct the robot to aim the coil during TMS delivery.
Using the TMS coil aiming parameters and the treatment site configuration (location, column orientation), the induced E field may be modeled using on-axis data to predict its magnitude at the delivery site. A fully 3-D approach, including effects of the patient head may be modeled for each subject using EMAS (Ansoft, Pittsburgh, Pa.) and anatomical data derived from CT scans.
In still further embodiments, methods for accelerating calculations of electric fields may be provided. EMAS is a general purpose Maxwell's equations solver. Because of its generality, it is slow and uses a lot of computing resources. Calculation of the electric field for a single coil position may require several hours for interactive, image-guided imaging abbreviation of this computation time is desirable. Computations may be accelerated chiefly by introduction of simplifying assumptions. For example, because the skull is a poor conductor, the magnetic fields produced by induced eddy currents can be safely ignored (Davey et al, 1991). As a consequence, the magnetic field produced by the TMS coil may be modeled as being unaffected by the head. Further, the skull is 100 times less conductive than soft-tissue. Therefore, it may be possible to treat the skull as a perfect insulator; treating all regions outside the skull as air. In addition, when using a standardized head model, it may be possible to pre-compute (and store as look-up tables) the E-field for often-stimulated brain zones, such as M1-hand. The E-fields for an individual could be then be calculated as a perturbation of the pre-computed values. Finally, it is possible that field estimates based just on the local skull curvature may be quite accurate and provide considerable computational savings.
To support safe movement of the coil about the head, 3-D models of the head and TMS coil surfaces may be defined, and a polygonal model of the surface of the TMS coil created. Full-surface polygonal models of each subject's head may be created from the high-resolution 3-D MR image. By thresholding the MR image to the head surface, a mask defining the full 3-D volume of the subjects head can be obtained. Using 3-D morphological dilation, a second volume that is guaranteed to be a minimum distance from the head surface can be created. The dilated head volume may be used to model a volume with a realistic safety margin for use in collision avoidance control when moving from point to point about the head. The undilated model may serve to determine closest approach without contact in final positioning. The surfaces of the head models may be extracted and stored using the marching cubes algorithm and its standard storage format (Schroeder et. al., 1996). Additionally or alternatively, collision avoidance may be provided by the operator (run/stop button with absolute shut-off switch).
In an example embodiment, TMS delivery software may be used to position the TMS coil exactly as prescribed by the planning software. This software may provide accurate robot-controlled positioning of the TMS coil along and about its z-axis. Additionally, movement and pose strategy may ensure safe, effective operation. Finally, feedback to the delivery software may monitor this operation.
In exemplary software, a graphical user interface may be provided, in which the user is prompted for input and presented with options for delivery of TMS. The TMS delivery may proceed and display status information. A graphical user interface may monitor movement of the robot/TMS coil about the patient's head. Polygonal models developed for collision avoidance may be used for rendering 3-D views of the head and TMS coil body. A 3-D polygonal model of the robot with TMS coil mounted may be used to create a real-time 3-D display of the robot moving about the head surface. View orientation may be user-selectable to provide unobstructed viewing of the patient's head. This display feature may be provided as animation during treatment planning to simulate delivery.
For a given TMS coil location and orientation, several arm poses will often be possible. Environmental factors, including TMS coil electrical and cooling lines, PET scanner housing, patient table and head holder, etc., may be accommodated when selecting best pose. Joint configuration may be under control of the supervisor software developed to move the arm, which are computed using the forward and inverse kinematics equations of the robot. The user may be presented with several options and asked to select one for the treatment delivery.
In an example embodiment, the robot can position the TMS coil but cannot rotate it about coil's z-axis, therefore the coil angle may be adjusted manually. Software may use the manual digitizer to measure the TMS coil z-axis rotation angle. In embodiments with a rotational sensor, the sensor may be used for coil-angle sensing, in preference to the manual digitizer. The current and target angle, and the difference may be actively displayed. Manual orientation may be done with TMS housing away from the head at the position where the robot has already aligned the coil along its z-axis. The z-axis rotation angle sensing may be automated using the precision rotary potentiometer. The sensor may be connected to the robot controller via its existing wiring harness. Movement toward the head along this treatment axis may be done after adjusting coil angle.
The cortical column orientation throughout the treatment volume may be determined by a method using both inner and outer cortical surfaces. The first step may be a semi-automated extraction of the cortical surface, using, for example, the MEDx (Version 2.1, Sensor Systems, Inc.) image processing software. This step requires approximately 20 minutes per surface. The extracted surface may be refined using morphologic operators (dilation, erosion, and 3-D connectivity routines), also within MEDx. Both inner and outer cortical surfaces for the treatment volume, including all visible sulci, may be extracted and used to create a binary mask of the cortical columns within the treatment volume. The binary mask may be used to create a set of normal vectors that uniformly fill each voxel within the mask. A 3-D Euclidean Distance Map (EDM) (Russ 1996, Gonzales and Wood 1992) may be used to fill the binary mask with distances from the outer to the inner surface. The EDM may then be smoothed with a 3-D Gaussian filter to produce a smooth gradient of values from the outer to inner cortical surface bounds. A normal vector may then be calculated for each voxel within the binary mask from the gradient of EDM values (Schroeder, et. al. 1996). These normal vectors may be converted to unit normal vectors. The sense of the unit normal vectors may be to point into the brain (orthodromic column direction). This method, based on the EDM provides a smooth and continuous orientation change for the normal vectors. These normal vectors may be used for the orthodromic orientation of the cortical columns for each voxel in the binary mask.
A 3-D scalar map of columnar component of E-field may be accomplished using 3-D models of the E field accounting for the head a set of coil aiming parameters (position, coil z-axis orientation, and coil rotation), and the 3-D model of cortical column orientations. Using the TMS coil's treatment parameters to orient the E-fields relative to the head, the scalar product of E-field vectors and cortical-column unit vectors can be calculated, yielding E-field magnitude along cortical columns (Volts/cm) for each voxel within the treatment volume. The result will be a 3-D scalar map of the columnar component of the E-field throughout the treatment volume. This map may be used to overlay onto the MR image to see the planned treatment. Modeled treatment effects may be visualized as a pseudo-color overlay on the anatomical image. An option may be provided to display the absolute value of the E-field magnitude, and a method to enable/disable the overlays. The user may adjust TMS coil global E-field magnitude and a threshold value for display to see the effect of these on target and non-target cortical regions. Since the E-field scalar map depends on both cortical column and coil orientation, a “jog” feature to interactively reposition the TMS coil may be provided to help the user optimize the magnitude in a targeted region. This jog feature allows small increments of movement along and rotations about the orthogonal tool axes after they have been verified safe (non-interfering with the avoidance volume) by the supervisory controller.
In certain embodiments, the metal in the PET gantry might alter the TMS-induced B and E fields. If so, this effect must be included in the E-field modeling. Conversely, the TMS B-field might affect the performance of the PET photo-multiplier tubes, which may be resolved with μ metal shielding during TMS/PET. Further, the TMS coil and coil holder may excessively, regionally attenuate the emitted annihilation photons. It is to be understood that standard attenuation corrections may suffice to correct for the effect of the TMS coil.
At step 430, the defined site may be registered to the subject's anatomy. In an example embodiment, the registration step may be performed via use of a 3-D MRI. At step 440, planning tasks may be performed. These planning tasks may include modeling various elements in order to optimize TMS delivery. These models may include, for example, a 3-D cortical model of the subject's brain, to determine the accurate location of cortical columns and cortex regions. In addition, a 3-D E-field may be modeled for the chosen TMS stimulator design. Further, in an example embodiment, a 3-D induction model may be performed also. The 3-D induction model is derived from the scalar product of the E-field and the cortical column direction, and is the net volts/cm estimated for cortical columns. Based on all this information, at step 450, the TMS treatment plan may be computed. In an example embodiment, the plan may include information regarding optimal coil position, coil orientation, power setting, rate/duration, and a 3-D head model. The TMS delivery plan developed at step 450 may then be provided to a TMS treatment delivery system. It is to be understood that the TMS treatment planning system described herein may be performed using a standard personal computer appropriately programmed, or it may be performed via specialized computers, such as a UNIX workstation or a mainframe computer.
In conjunction with the CCCAP disclosed herein, a higher degree of control over the electrical field induced by a TMS coil will improve the value of the experiment/treatment. Such control may be aided by new and inventive coil designs.
Greater control may be had over such features as coil inductance and power dissipation with methods that have been in use for designing gradient coils for magnetic resonance imaging (“MRI”)since the late 1980's for cylindrical coils (Turner) and later for planar coils (Martens et al.). These methods involve an inverse technique which allows the magnetic field to be defined at certain points in a volume. This method may be adapted to the design of magnetic stimulation coils which enables the electric field to be defined at certain points. While it is not reasonable to focus the magnetic field at an arbitrary position inside the brain/muscle volume, the resultant coils may provide more focused fields in two dimensions near the brain/muscle surface and with other desirable characteristics such as a lower inductance or a lower heat dissipation.
Suggestions of 10 kHz as the optimum frequency for the stimulation waveform have been made (Davey and Epstein). The stimulation circuit may be considered primarily an LC circuit with a small resistive loss. The capacitive section of the circuit stores energy and when a switch is closed, the energy is transferred to the inductive section, the magnetic stimulation coil, and back to the capacitive section with a frequency ω=1/(LC)˝. The voltage induced in a conducting body is proportional to this frequency. For the purposes of coil design, the coil is of interest, and thus the capacitive section may be considered constant. Hence, to increase the resonant frequency and the induced voltage, the coil inductance may be lowered.
The target field method may be adapted to provide a coil with a minimum inductance for a given set of electric field constraints. The constraints may be set to provide as smooth a design as possible. The constraints should also be set such the coil is of a reasonable size in order that the natural field falloff moving away from the coil is not too sharp.
The theory behind minimum inductance TMS coils may be adapted from known design techniques for target field gradient coil designs for MRI, such as that of Martens et al, and Turner. For current flowing on a planar surface, the x-y plane, the Fourier transform pair may be defined as:
j x(α, β)=∫∫e −iαx e −iβy J x(x, y)dxdz (1.1)
and similarly for the y-component of the current density. Using:
div J=0 (3)
leads to, at the plane y=a:
αj x(α, β)+βj z(α, β)=0 (4)
From the vector potential:
Orthogonal components may be determined:
A z(r)=0. (8)
The electric field in a homogeneous media may be written as:
and so for a sinusoidal current, I=I0 sin ωt,
E x(r)=A x(r)ω cos ωt etc. (10)
If the x-component of the E-field is considered:
|E x |=ω.A x (11)
The electric field must be symmetric in both the x and y directions, so that the imaginary components in the above equation can be reduced to cos(αx) and cos(βy). The energy from the current density plane may be calculated from:
In order to minimize W subject to a set of electric field constraints, the error function may be formed:
where λj are Lagrange multipliers and the electric field at points rj is constrained to have the values Et(rj). Setting:
The Lagrange multipliers are found by solving the matrix equation:
found by substituting the current density equation into the electric field equation. Once the current density has been calculated, the Fourier transform may be used to obtain the current density Jx(x,z) and Jz(x,z). These may then be integrated to provide a stream function and may be discretized into current loops. The inductance may be estimated using W=0.5*L.I2. The accuracy of the design is then dependent on the number of loops chosen to approximate the continuous current density. The design can then be checked using:
and summing individual wire elements to obtain the total field distribution. The coil may also be evaluated in the presence of conducting boundaries to simulate a human head using numerical methods such as finite element analysis. The design process does not take into account these boundaries, although an approximation may be incorporated into the design using a conducting sphere and a well-known method of calculating electric field in a spherical volume, as set forth in Eaton (1992).
To obtain a varying electric field, the symmetries may be adjusted, i.e. all cos(αx) need to be changed to sin(αx).
To achieve a minimum power design, the above calculation may be used, but replacing the energy term with the following:
Recent studies have suggested that rapid stimulation (>1 Hz) may enhance the potency of magnetic stimulation. For situations involving rapid stimulation, heating of the coil becomes an issue due to the increased duty cycle. While cooling mechanisms have been employed it is also advantageous to include in the coil design a power dissipation term, in order for heating to be minimized. Minimum power designs are larger and generally smoother than minimum inductance designs.
When attempting TMS, unwanted stimulation of the nerves in the scalp may occur for persons with high thresholds (usually meaning a slightly thicker skull), possibly causing pain and facial twitching, confounding experimental results. With the addition of negative turns to the windings of the stimulator coil, and a small spacing between the coil and the scalp, the fields across this area may be reduced significantly. However, this comes at the expense of some focusing ability, and coil inductance and heating.
The spacing between windings 640 may vary in given embodiments. As shown in
The coil may be designed in accordance with standard practices for coil manufacture well-known to those with skill in the art. In an example embodiment, the windings may comprise copper or another conducting material, such as silver (or other conducting material) ribbon placed on its edge. Such wire may have a diameter of between about 0.1 and 1.0 millimeters in example embodiments. Furthermore, the coil may be encased in plastic or another non-conducting material in order for reasons of safety and other issues. In certain embodiments, the coil may be encased in thermally conductive epoxy to enhance heat dissipation. Further, the coil may be encased with a water, oil or air cooling system. Coils may also be wound in layers placed directly upon one another, and connected in series or parallel. The wire pattern may be transferred to a sheet former for construction using a computerized milling machine, or may be transferred using heat from an inked hard copy, such as a computer printout. The wire pattern may be constructed from one single wire/ribbon wound with small connecting paths between loops. Standard connections from the coil to cabling necessary to adapt the coil to a magnetic stimulator may be made.
The method may be changed slightly to provide a minimum inductance configuration providing a spatial gradient in the electric field for peripheral nerve stimulation. An example of this is shown in
In certain embodiments, it may be desirable for the coil to be designed in a generally square manner. The coil shown at
The constraints in this embodiment were set such that the x-component of the electric field at the center point was twice that of a point at 2.5 cm along the x and y axes and 3 cm from the coil. That is, the constraints were:
x(cm) y z E-field (V/m) 0.0 0.0 3.0 100 2.5 1.1 3.0 50 0.0 2.5 3.0 50
It is to be understood that in certain embodiments, the negative turns may be located in a separate layer from turns having an opposite polarity. For example, small secondary coil, consisting of a few turns of either a figure 8 design, or a smaller version of the target field designs disclosed herein, may be placed next to the magnetic stimulation coil such that the center of the coils are coincident but such that the distance between them is adjustable. The current may flow in the second coil such that the electric field generated by the current pulse is opposing the direction of the field produced by the main coil. Further, the angle that the loops make to the surface of the coil may also be adjusted.
It is to be understood that the coil designs disclosed herein may be used independently from the delivery methods herein. Further, it is to be understood that the delivery methods disclosed may be used independently from the coils herein.
While particular embodiments of the present invention have been shown and described, it will be obvious to those skilled in the art that, based upon the teachings herein, changes and modifications may be made without departing from this invention and its broader aspects and, therefore, the appended claims are to encompass within their scope all such changes and modifications as are within the true spirit and scope of this invention. Furthermore, it is to be understood that the invention is solely defined by the appended claims.
The following references, to the extent that they provide exemplary procedural or other details supplementary to those set forth herein, are specifically incorporated herein by reference.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4864966 *||Feb 5, 1988||Sep 12, 1989||Automated Artists Corp.||Robotic airbrush apparatus|
|US4937759 *||Jun 15, 1988||Jun 26, 1990||Robotics Research Corporation||Industrial robot with controller|
|US4940453||Jan 28, 1987||Jul 10, 1990||Cadwell Industries, Inc.||Method and apparatus for magnetically stimulating neurons|
|US4990839 *||Dec 9, 1988||Feb 5, 1991||Schonlau William J||Modular robotic system|
|US5047005 *||Jul 5, 1990||Sep 10, 1991||Cadwell Industries, Inc.||Method and apparatus for magnetically stimulating neurons|
|US5078140 *||Sep 23, 1986||Jan 7, 1992||Kwoh Yik S||Imaging device - aided robotic stereotaxis system|
|US5116304||May 10, 1991||May 26, 1992||Cadwell Industries, Inc.||Magnetic stimulator with skullcap-shaped coil|
|US5236432||Aug 24, 1992||Aug 17, 1993||Board Of Regents Of The University Of Washington||Robot-aided system for surgery|
|US5427097 *||Dec 10, 1992||Jun 27, 1995||Accuray, Inc.||Apparatus for and method of carrying out stereotaxic radiosurgery and radiotherapy|
|US5466213 *||Jan 6, 1994||Nov 14, 1995||Massachusetts Institute Of Technology||Interactive robotic therapist|
|US5697285 *||Dec 21, 1995||Dec 16, 1997||Nappi; Bruce||Actuators for simulating muscle activity in robotics|
|US5725471||Nov 28, 1994||Mar 10, 1998||Neotonus, Inc.||Magnetic nerve stimulator for exciting peripheral nerves|
|US5806518||Sep 11, 1995||Sep 15, 1998||Integrated Surgical Systems||Method and system for positioning surgical robot|
|US6013997 *||Aug 7, 1998||Jan 11, 2000||Tower Automotive, Inc.||Three dimensional tactile seam tracing device|
|US6035228 *||Nov 28, 1997||Mar 7, 2000||Picker International, Inc.||Frameless stereotactic arm apparatus and method of using same|
|US6048302 *||Oct 16, 1998||Apr 11, 2000||Bio-Magnetic Therapy Systems, Inc.||Apparatus for the treatment of disorders of tissue and/or the joints|
|US6076008||Feb 3, 1999||Jun 13, 2000||St. Louis University||System for indicating the position of a surgical probe within a head on an image of the head|
|US6086525||Dec 31, 1997||Jul 11, 2000||Neotonus, Inc.||Magnetic nerve stimulator for exciting peripheral nerves|
|US6099459 *||Sep 4, 1998||Aug 8, 2000||Jacobson; Jerry I.||Magnetic field generating device and method of generating and applying a magnetic field for treatment of specified conditions|
|US6132361||Aug 15, 1997||Oct 17, 2000||Neotonus, Inc.||Transcranial brain stimulation|
|US6179770||Apr 16, 1999||Jan 30, 2001||Magstim Company Limited||Coil assemblies for magnetic stimulators|
|US6179771||Apr 21, 1999||Jan 30, 2001||Siemens Aktiengesellschaft||Coil arrangement for transcranial magnetic stimulation|
|US6198958||Jun 11, 1998||Mar 6, 2001||Beth Israel Deaconess Medical Center, Inc.||Method and apparatus for monitoring a magnetic resonance image during transcranial magnetic stimulation|
|US6212419 *||Nov 10, 1998||Apr 3, 2001||Walter M. Blume||Method and apparatus using shaped field of repositionable magnet to guide implant|
|US6266556 *||Apr 27, 1998||Jul 24, 2001||Beth Israel Deaconess Medical Center, Inc.||Method and apparatus for recording an electroencephalogram during transcranial magnetic stimulation|
|US6330467 *||Apr 6, 1999||Dec 11, 2001||Stereotaxis, Inc.||Efficient magnet system for magnetically-assisted surgery|
|US6366814||Dec 29, 2000||Apr 2, 2002||Birinder R. Boveja||External stimulator for adjunct (add-on) treatment for neurological, neuropsychiatric, and urological disorders|
|US6459924 *||Nov 10, 1998||Oct 1, 2002||Stereotaxis, Inc.||Articulated magnetic guidance systems and devices and methods for using same for magnetically-assisted surgery|
|US6585746||Jan 30, 2001||Jul 1, 2003||Philip L. Gildenberg||Hair transplantation method and apparatus|
|US6783524 *||Apr 18, 2002||Aug 31, 2004||Intuitive Surgical, Inc.||Robotic surgical tool with ultrasound cauterizing and cutting instrument|
|DE19909587A1||Mar 4, 1999||Feb 10, 2000||Pulse Medical Co||Electromagnet for a magnet-type medical instrument which can kill a pain in a human body|
|GB2261820A||Title not available|
|1||Boroojerdi et al., Localization of the Motor Hand Area Using Transcranial Magnetic Stimulation and Functional Magnetic Resonance Imaging (Apr. 1999); Clin. Neurophysiol vol. 110, No. 4: pp. 699-704.|
|2||Caldwell, Optimizing Magnetic Stimulator Design (1991); Magnetic Motor Stimulation: Basic Principles and Clinical Experience (EEG Supp.. 43): pp. 238-248.|
|3||Davey, Magnetic Stimulation Coil and Circuit Design (Nov. 2000); IEEE Transactions on Biomedical Engineering, vol. 47, No. 11: pp. 1493-1499.|
|4||Guggisberg et al., Motor evoked potentials from masseter muscle induced by Transcranial magnetic stimulation of the pyramidal tract: the importance of coil orientation (Dec. 2001), Clin Neurophysiol. 2001; 112(12), pp. 2312-2319.|
|5||Herwig et al., The navigation to transcranial magnetic stimulation (Nov. 30, 2001), Psychiatry Res.; 108(2); pp. 123-131.|
|6||Hsu et al., A 3-D Differential Coil Design for Localized Magnetic Stimulation (Oct. 2001); IEEE Transactions on Biomedical Engineering, vol. 48, No. 10: pp. 1162-1168.|
|7||Ilmoniemi et al., "Transcranial Magnetic Stimulation-A New Tool for Functional Imaging of the Brain". Critical Reviews in Biomedical Engineering, CRC Press. vol. 27, No. 3-5, 1999, pp. 241-284.|
|8||Krings et al., Stereotactic Transcranial Magnetic Stimulation: Correlation with Direct Electrical Cortical Stimulation (Dec. 1997); Neurosurgery, vol. 41, No. 6: pp. 1319-1326.|
|9||Martens et al., Insertable Biplanar Gradient Coils for Magnetic Resonance Imaging (1991); American Institute of Physics, Rev. Sci. Instrum. vol. 62, No. 11: pp. 2639-2645.|
|10||Mouchawar et al., Guidelines for Energy-Efficient Coils: Coils Designed for Magnetic Stimulation of the Heart (1991); Magnetic Motor Stimulation: Basic Principles and Clinical Experience (EEG Supp.. 43): pp. 255-267.|
|11||Narayama et al. "Use of Neurosurgical Robot for Aiming and Holding in Cortical TMS Experiments", NEUROIMAGE. vol. 11, No. 5, 2000. p. S471.|
|12||NEUROMATE-Integrated Surgical Systems (http://www.robodoc.com/eng/neuromate.html).|
|13||Paus et al., Transcranial Magnetic Stimulation During PET: Reaching and Verifying the Target Site (1998); Hum. Brain Map, vol. 6, No. 5-6: pp. 399-402.|
|14||Paus et al., Transcranial Magnetic Stimulation during Positron Emission Tomography: A New Method for Studying Connectivity of the Human Cerebral Cortex (May 1, 1997); The Journal of Neuroscience, vol. 17, No. 9: pp. 3178-3184.|
|15||Paus, Imaging the Brain Before, During, and After Transcranial Magnetic Stimulation (Feb. 1999); Neuropsychologia, vol. 37, No. 2: pp.. 219-224.|
|16||PCT/US02/14157 International Preliminary Examination Report Mailed Jul. 8, 2003.|
|17||PCT/US02/14157 International Search Report Mailed Apr. 7, 2003.|
|18||Ruohonen et al., Coil Design for Real and Sham Transcranial Magnetic Stimulation (Jan. 2000); IEEE Transactions on Biomedical Engineering, vol. 47, No. 2. pp. 145-148.|
|19||Ruohonen et al., Functional Magnetic Stimulation: Theory and Coil Optimization (1998); Bioelectrochemistry and Bioenergietics 47: pp. 213-219.|
|20||Supplementary Partial European Search Report for EPO Application No. 02724022.5-2305; Mailed Jul. 11, 2005.|
|21||Supplementary Partial European Search Report for EPO Application No. 02724022.5-2305; Mailed Nov. 10, 2005.|
|22||Traad, Monique. "A Quantitative Positioning Device For Transcranial Magnetic Stimulation". Engineering In Medicine and biology Society, 1990. Proceedings of the Twelfth Annual International Conference of the IEEE. Philadelphia, PA, Nov. 1-4, 1990. p. 2246.|
|23||Turner, A Target Field Approach to Optimal Coil Design (1986); The Institute of Physics 19: L147-L151.|
|24||Turner, Minimum Inductance Coils (1988); IOP Publishing Ltd.: pp. 948-952.|
|25||Wasserman et al., Locating the Motor Cortex on the MRI with Transcranial Magnetic Stimulation (Feb. 1996); NeuroImage, vol. 3, No. 1: pp. 1-9.|
|26||Weh-Hau Lin et al., Magnetic Coil Design Considerations for Functional Magnetic Stimulation (May 2000); IEEE Transactions on Biomedical Engineering, vol. 47, No. 5: pp. 600-610.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7539528 *||Sep 18, 2003||May 26, 2009||Jinhu Xiong||Using magnetic resonance imaging to directly map neuronal activity|
|US7720519 *||Oct 16, 2003||May 18, 2010||Elekta Neuromag Oy||Method for three-dimensional modeling of the skull and internal structures thereof|
|US7753836||Mar 28, 2007||Jul 13, 2010||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US7925066 *||Sep 11, 2007||Apr 12, 2011||Nexstim Oy||Method and apparatus for correcting an error in the co-registration of coordinate systems used to represent objects displayed during navigated brain stimulation|
|US7946973||Nov 6, 2009||May 24, 2011||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US8018232 *||Mar 31, 2009||Sep 13, 2011||General Electric Company||Interleaved gradient coil for magnetic resonance imaging|
|US8179135 *||Jan 26, 2009||May 15, 2012||California Institute Of Technology||Low field electron paramagnetic resonance imaging with SQUID detection|
|US8265910 *||Oct 9, 2008||Sep 11, 2012||Cervel Neurotech, Inc.||Display of modeled magnetic fields|
|US8267850||Nov 26, 2008||Sep 18, 2012||Cervel Neurotech, Inc.||Transcranial magnet stimulation of deep brain targets|
|US8303480||Sep 28, 2011||Nov 6, 2012||Mclean Hospital Corporation||Magnetic field treatment techniques|
|US8523753||Aug 15, 2012||Sep 3, 2013||Cervel Neurotech, Inc.||Transcranial magnet stimulation of deep brain targets|
|US8543189||Apr 17, 2008||Sep 24, 2013||Medtronic Navigation, Inc.||Method and apparatus for electromagnetic navigation of a magnetic stimulation probe|
|US8545378||May 31, 2007||Oct 1, 2013||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US8702582||Nov 5, 2012||Apr 22, 2014||Mclean Hospital||Magnetic field treatment techniques|
|US8723628||Jan 7, 2010||May 13, 2014||Cervel Neurotech, Inc.||Shaped coils for transcranial magnetic stimulation|
|US8795148||Oct 26, 2010||Aug 5, 2014||Cervel Neurotech, Inc.||Sub-motor-threshold stimulation of deep brain targets using transcranial magnetic stimulation|
|US8845508||Mar 11, 2009||Sep 30, 2014||The Board Of Trustees Of The Leland Stanford Junior University||Robotic apparatus for targeting and producing deep, focused transcranial magnetic stimulation|
|US8956273||Aug 20, 2008||Feb 17, 2015||Cervel Neurotech, Inc.||Firing patterns for deep brain transcranial magnetic stimulation|
|US8956274||Jul 16, 2010||Feb 17, 2015||Cervel Neurotech, Inc.||Transcranial magnetic stimulation field shaping|
|US9132277||Apr 7, 2014||Sep 15, 2015||Cerval Neurotech, Inc.||Shaped coils for transcranial magnetic stimulation|
|US9132278||Mar 12, 2014||Sep 15, 2015||Brainsway, Ltd.||Transcranial magnetic stimulation system and methods|
|US9248308||Feb 21, 2013||Feb 2, 2016||Brainsway, Ltd.||Circular coils for deep transcranial magnetic stimulation|
|US9254394||Feb 21, 2013||Feb 9, 2016||Brainsway, Ltd.||Central base coils for deep transcranial magnetic stimulation|
|US9256951 *||Nov 17, 2010||Feb 9, 2016||Koninklijke Philips N.V.||System for rapid and accurate quantitative assessment of traumatic brain injury|
|US9265965||Sep 27, 2012||Feb 23, 2016||Board Of Regents, The University Of Texas System||Apparatus and method for delivery of transcranial magnetic stimulation using biological feedback to a robotic arm|
|US9345901||May 19, 2010||May 24, 2016||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US9352167||Jun 27, 2011||May 31, 2016||Rio Grande Neurosciences, Inc.||Enhanced spatial summation for deep-brain transcranial magnetic stimulation|
|US9381374||Sep 14, 2015||Jul 5, 2016||Rio Grande Neurosciences, Inc.||Shaped coils for transcranial magnetic stimulation|
|US9486639||Sep 22, 2011||Nov 8, 2016||The Board Of Trustees Of The Leland Stanford Junior University||Trajectory-based deep-brain stereotactic transcranial magnetic stimulation|
|US9492679||Jul 15, 2011||Nov 15, 2016||Rio Grande Neurosciences, Inc.||Transcranial magnetic stimulation for altering susceptibility of tissue to pharmaceuticals and radiation|
|US9504846||Nov 13, 2013||Nov 29, 2016||Mehran TALEBINEJAD||Circuit and method for use in transcranial magnetic stimulation|
|US9533168||Feb 21, 2013||Jan 3, 2017||Brainsway, Ltd.||Unilateral coils for deep transcranial magnetic stimulation|
|US20040096395 *||Sep 18, 2003||May 20, 2004||Jinhu Xiong||Using magnetic resonance imaging to directly map neuronal activity|
|US20050215889 *||Mar 29, 2004||Sep 29, 2005||The Board of Supervisory of Louisiana State University||Methods for using pet measured metabolism to determine cognitive impairment|
|US20050256539 *||Mar 25, 2003||Nov 17, 2005||George Mark S||Methods and systems for using transcranial magnetic stimulation to enhance cognitive performance|
|US20060052687 *||Oct 16, 2003||Mar 9, 2006||Jarmo Ruohonen||Method for three-dimensional modeling of the skull and internal structures thereof|
|US20060122496 *||Nov 17, 2004||Jun 8, 2006||Mark George||Method, apparatus, and system for automatically positioning a probe or sensor|
|US20060173274 *||Jul 11, 2003||Aug 3, 2006||George Mark S||Functional magnetic resonance imaging guided transcranial magnetic stimulation deception inhibitor|
|US20070293916 *||Mar 28, 2007||Dec 20, 2007||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US20080064950 *||Sep 11, 2007||Mar 13, 2008||Jarmo Ruohonen||Method and Apparatus for Correcting an Error in the Co-Registration of Coordinate Systems Used to Represent Objects Displayed During Navigated Brain Stimulation|
|US20080262338 *||Apr 17, 2008||Oct 23, 2008||Medtronic Navigation, Inc||Method and apparatus for electromagnetic navigation of a magnetic stimulation probe|
|US20090099405 *||Aug 4, 2008||Apr 16, 2009||Neostim, Inc.||Monophasic multi-coil arrays for trancranial magnetic stimulation|
|US20090156884 *||Nov 26, 2008||Jun 18, 2009||Schneider M Bret||Transcranial magnet stimulation of deep brain targets|
|US20090227830 *||Mar 10, 2008||Sep 10, 2009||Neuronetics, Inc.||Apparatus for coil positioning for tms studies|
|US20090234243 *||Mar 11, 2009||Sep 17, 2009||Schneider M Bret||Robotic apparatus for targeting and producing deep, focused transcranial magnetic stimulation|
|US20090295390 *||Jan 26, 2009||Dec 3, 2009||California Institute Of Technology||Low field electron paramagnetic resonance imaging with squid detection|
|US20100087699 *||Nov 6, 2009||Apr 8, 2010||The Trustees Of Columbia University In The City Of New York||Systems and methods for inducing electric field pulses in a body organ|
|US20100113863 *||Oct 28, 2009||May 6, 2010||Medical University Of South Carolina||Methods and Systems for Using Transcranial Magnetic Stimulation to Enhance Cognitive Performance|
|US20100185041 *||Jan 20, 2009||Jul 22, 2010||Lee Richard H||Cylindrical magnetic field projection method|
|US20100185042 *||Feb 5, 2010||Jul 22, 2010||Schneider M Bret||Control and coordination of transcranial magnetic stimulation electromagnets for modulation of deep brain targets|
|US20100244836 *||Mar 31, 2009||Sep 30, 2010||General Electric Company||Interleaved gradient coil for magnetic resonance imaging|
|US20100256438 *||Aug 20, 2008||Oct 7, 2010||Mishelevich David J||Firing patterns for deep brain transcranial magnetic stimulation|
|US20100256439 *||Aug 12, 2008||Oct 7, 2010||Schneider M Bret||Gantry and switches for position-based triggering of tms pulses in moving coils|
|US20100286468 *||Oct 27, 2008||Nov 11, 2010||David J Mishelevich||Transcranial magnetic stimulation with protection of magnet-adjacent structures|
|US20100286470 *||Jul 16, 2010||Nov 11, 2010||Schneider M Bret||Transcranial magnetic stimulation field shaping|
|US20100298623 *||Oct 24, 2008||Nov 25, 2010||Mishelevich David J||Intra-session control of transcranial magnetic stimulation|
|US20100331602 *||Sep 9, 2008||Dec 30, 2010||Mishelevich David J||Focused magnetic fields|
|US20110004450 *||Oct 9, 2008||Jan 6, 2011||Mishelevich David J||Display of modeled magnetic fields|
|US20110098779 *||Oct 26, 2010||Apr 28, 2011||Schneider M Bret||Sub-motor-threshold stimulation of deep brain targets using transcranial magnetic stimulation|
|US20110207988 *||Oct 31, 2008||Aug 25, 2011||Nexstim Oy||Method, apparatus and computer program for non-invasive brain stimulation when target muscles are suitably active|
|US20110270345 *||Aug 11, 2008||Nov 3, 2011||Johns Hopkins University||Focal noninvasive stimulation of the sensory cortex of a subject with cerebral palsy|
|US20120327075 *||Nov 17, 2010||Dec 27, 2012||Trustees Of Dartmouth College||System for rapid and accurate quantitative assessment of traumatic brain injury|
|WO2009023162A1 *||Aug 11, 2008||Feb 19, 2009||The Johns Hopkins University||Focal noninvasive stimulation of the sensory cortex of a subject with cerebral palsy|
|WO2009023680A1 *||Aug 12, 2008||Feb 19, 2009||Neostim, Inc.||Gantry and switches for position-based triggering of tms pulses in moving coils|
|International Classification||A61N2/04, A61N2/02, A61N1/00|
|Jul 30, 2002||AS||Assignment|
Owner name: BOARD OF REGENTS, THE UNIVERSITY OF TEXAS SYSTEM,
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:FOX, PETER;LANCASTER, JACK;DODD, STEPHEN;REEL/FRAME:013133/0437;SIGNING DATES FROM 20020708 TO 20020709
|May 2, 2008||AS||Assignment|
Owner name: NATIONAL INSTITUTES OF HEALTH (NIH), U.S. DEPT. OF
Free format text: EXECUTIVE ORDER 9424, CONFIRMATORY LICENSE;ASSIGNOR:UNIVERSITY OF TEXAS HEALTH SCIENCE CENTER AT SAN ANTONIO;REEL/FRAME:020892/0590
Effective date: 20020722
|Jan 22, 2010||FPAY||Fee payment|
Year of fee payment: 4
|Jan 28, 2014||FPAY||Fee payment|
Year of fee payment: 8